KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT

WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

How can we improve mathematical The STeP Journal

vocabulary comprehension that will Student Teacher Perspectives
allow students to develop higher-order Copyright © 2016
levels of learning? University of Cumbria
Vol 3 (2) pages 47-61

Vivienne Kenyon
Cornwall School Centred Initial Teacher Training

Abstract
The style of the mathematics GCSE examinations in England is changing under the new
reform (Department for Education, 2013a), which promises to test in problem-solving and
worded contexts to a greater degree. This will undoubtedly present a barrier for many
students due to the increased need to understand and use mathematical vocabulary. During
this enquiry, I aimed to find a teaching strategy that would help students to learn
mathematical vocabulary, and so enabling them to access higher-order levels of learning in
preparation for the new exams. The strategy used was the Frayer Model, implemented for a
6-week period with a class of 15 participants. Results from pre- and post-instruction testing
were analysed, along with the results from a control class. The class who received the
instruction showed clear improvements over the control class, however the model appeared
to be less effective with those participants with a lower reading age. Also, the
questionnaires highlighted that those who were more engaged with the model generally
experienced greater improvements in their understanding of mathematical vocabulary in
the test.

Introduction
One of the most significant reforms in the education system in England in recent times is the
introduction of the new GCSE examinations (Department for Education, 2013a). The new
mathematics summative assessments will be rolled out next year, and there is currently
much anticipation and uncertainty regarding their style and complexity. However, we do
know is that they will examine mathematics in a much more applied context with a much
greater need for students to read and understand worded problems that will test not only
their knowledge and skills, but also their ability to reason and present robust mathematical
arguments (Department for Education, 2013a). Mathematical vocabulary has always been
an important yet somewhat over-looked aspect of learning mathematics that perhaps has
previously not always received the attention that it deserves (Adams, 2003). But now there
can be no hiding from the fact that it can pose as a significant barrier to learning and
achievement in the imminent reformed examinations. So how can we improve
mathematical vocabulary comprehension that will allow students to develop these higher-
order levels of learning?

Literature Review
There is known to be a strong correlation between vocabulary knowledge and reading
comprehension ability (Stanovich, 2008), and it follows that reading comprehension is
essential in learning how to problem solve effectively (Carter, 2006; Amen, 2006), a
fundamental part of the new assessment strategy. Marzano (2004) specifically linked
vocabulary knowledge to overall academic achievement, and precisely the same relationship

Citation
Kenyon, V. (2016) ‘How can we improve mathematical vocabulary comprehension that will
allow students to develop higher-order levels of learning?’, The STeP Journal, 3(2), pp.47-61.
47
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

has been found between mathematical vocabulary and mathematical performance

(Riccomini et al., 2015). For decades, it has been known that linguistic abilities have an
impact on mathematical performance (Aiken, 1972), and more recently, Morgan (1999) and
Schleppegrell (2007) emphasised how learning to understand and use mathematical
language is a critical part of learning mathematics. These literacy skills are so heavily
demanded in the new GCSE assessments, yet still do not seem to hold the weight that they
deserve in Maths lessons, and it appears that without these literacy skills, students are ill-
equipped to progress their learning to the standard required to achieve a pass grade.

Mathematics is deemed to be a language in itself, as it possesses all of the characteristics of

any other spoken language (Wakefield, 2000). Yet unlike most other languages that we are
familiar with, we do not read it merely from left to right, but also from right to left, top to
bottom, bottom to top, or a combination of all four directions (Adams, 2003). It has a
complex and precise structure, being composed of numerals and symbols as well as words.
This use of symbols immediately causes turmoil in the minds of some students, as there is
additional cognitive processing required to translate those symbols not only into words so
the mathematical sentence can be read, but also into meaning so the mathematical
sentence can be resolved.

Yet the mind cannot rest easy when reading mathematical words either; these are often
words that are unfamiliar, multi-syllabic, difficult to read and spell, and carry with them a
precise and sometimes highly abstract concept (Chinn, 2012); for example, the word
‘isosceles’ falls into this category. In addition to this, assessment questions are often
complicated by implied meanings associated with this type of terminology; therefore, if a
student is told that the shape in the question is a regular hexagon, they must instantly
realise that they may be required to draw upon any of the shape’s properties (such as
number of sides, symmetry, angle facts or tessellation) without being explicitly told, in order
to successfully find a solution (Thompson and Rubenstein, 2014).

There is a second set of mathematical words that students may be more familiar with and
would have experienced in everyday conversations (Gough, 2007), and one may be fooled
into thinking that this is a helpful circumstance. However, these words rarely transmit the
same meaning when placed in a mathematical context; for example, the word ‘volume’ falls
into this category. The original word meanings that have been formed and reinforced in the
brain over the years must now be altered in order to accommodate a new path to a
somewhat intangible meaning, potentially causing conceptual confusion.

Whether reading mathematical symbols or words, the route to success lies in the vocabulary
comprehension (Fisher and Frey, 2014), and hence it is important to all ability levels.
Bloom’s model of taxonomy (1956) and the SOLO taxonomy of Biggs and Collis (1982) both
indicate that learners must be able to identify, define and describe concepts (at the heart of
which is the content vocabulary) before progressing on to higher-level thinking and
reasoning. In addition to these learning classifications, the Subject Content and Assessment
Objectives for Mathematics (Department for Education, 2013b) specifically states that
students of all abilities will be tested on their vocabulary in particular content areas; for
example, both foundation level and higher level paper will require examinees to:
use the concepts and vocabulary of prime numbers, factors (divisors), multiples, common
factors, common multiples, highest common factor, lowest common multiple [and] prime
factorisation (Department for Education, 2013b)

48
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

When it comes to worded problems in assessments, there is much research that considers
that impact that mathematical vocabulary has on achievement. Ofqual (2015, p107) looked
specifically at the expected and actual difficulty levels of the new GCSE mathematics papers
from the various exam boards, and concluded that, beyond the maths involved, the way the
questions were worded and the context had an effect on the difficulty level. They also noted
that OCR1 examination board had a slightly higher word count which correlated with a
slightly higher level of difficulty. However, looking at word count alone is not a sophisticated
enough method to assess the impact of language in the questions.

Kan and Bulut (2015) found that changing the language used in mathematics questions
affected student performance in both directions: if numeric questions were converted into
worded questions, students’ performance worsened if the language became technical, but
improved if the language became colloquial. Either way, they highlighted a bias towards
those students who were more proficient readers, meaning that test scores could not be
compared on the grounds of pure mathematical ability. This is an interesting discussion
point – should we be testing pure mathematical ability, or is the language of mathematics an
intrinsic part of the learning that should simultaneously be tested? I believe that the
evidence here suggests that the language cannot be differentiated from the mathematics if
more than surface level learning is to be achieved.

Aims and Purposes

Background
From my previous classroom observations, I identified that vocabulary comprehension was a
particular weakness across all ability levels; in a middle ability set, the technical term
‘congruent’ caused confusion in one of the summative assessments, and in a top set class,
even the word ‘fully’ in the expression ‘fully factorise’ caused perplexity for some students.
As teachers, we may sometimes be inclined to frivolously hand over the definition in order
that the students get to the mathematical skills needed, but in a GCSE examination, this lack
of understanding could be the barrier that is the difference between a pass and a fail.

Aim
The aim of this enquiry was to find out if students were able to overcome a barrier to
learning, which took the form of mathematical vocabulary comprehension, by introducing a
focused teaching strategy.

Purposes
The purpose of the enquiry was to improve students’ overall mathematical ability in the
classroom and in assessments. I also carried out this enquiry to help to develop my own
teaching practice, and that of other teachers by disseminating my findings.

Ethical Considerations
Ethical considerations were at the forefront of my enquiry at all times, although this did
occasionally cause some conflict with my research objectives. My intention was to follow the
guidelines set out by the British Educational Research Association (BERA, 2011), which first
of all required that I inform and subsequently gain permission from the class teacher and
head of department to proceed with the enquiry.

1 http://www.ocr.org.uk/about-us/who-we-are/

49
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

Once permission was granted, the issue of consent from the participants arose. Cohen,
Manion and Morrison (1993, pp77-81) comment extensively on the issue of informed
consent of the participants, particularly of children, and that there are very few instances
where consent would not be deemed to be as important as other conflicting ethical factors.
This emerged as the main conflict to my research aims; my dilemma was whether to fully
inform the participants of these aims and risk damaging the integrity of the data. Barker
Bausell (2015, p66) writes about the Hawthorne Effect whereby participants may act out of
character if they know they are being watched for a particular reason, and part of my study
was to observe how the students involved themselves with the vocabulary model.

I had ownership of the class with whom I was intervening, and since the primary objective of
any teacher is to improve the ability of their students, I did not perceive the strategy that I
was using or my intentions to be deceiving the students in any way. However, there is a
difference between a teacher who performs action research in their class as a part of their
on-going aim to improve the learning of their students, and that of an action researcher who
plans to more formally present and disseminate their work, the latter creating the impetus
to abide by the formal ethical guidelines (Zeni, 1998). Therefore, to strike an acceptable
balance, I informed the students of just the broad context at the start of the action research
in order to allow them to give their informed consent to participate.

BERA (2011, section 16) states that children should be facilitated to give their fully informed
consent where possible, so having studied the school’s Record of Need, detailing SEN/D
(Special Educational Needs/Disabilities) and any other support that the children require, I
was confident that all participants in the study would be able to understand the simple
details of my enquiry, which would be sufficient for them to decide if they wanted to
participate. I gave all participants the opportunity to withdraw their work from the study if
they did not wish to continue to participate at any point.

In an effort to protect the anonymity of all participants (BERA, 2011, section 25) I have
removed all student and teacher names from the data collected, and will refer to students,
teachers and the school with numeric suffixes, if it is necessary to distinguish particular
participants. Similarly, I have made every effort to keep any personal information
confidential by sharing it only with those deemed to be necessary to the purposes of the
enquiry (BERA, 2011, section 25).

Methodology
Approach
The enquiry approach that I have used is that of action researcher. Bell and Waters (2014)
describe this approach as “applied research, carried out by practitioners who have
themselves identified a need for change or improvement”. It is a strong form of enquiry as it
allows the researcher to be part of the research, having the advantage that they know the
full purpose and methodology of the study, helping to improve its integrity and ethical
validity (Elton-Chalcraft, Hansen and Twiselton, 2008, p60).

Cohen, Manion and Morrison (1993, p7) discuss the two leading paradigms in the field of
educational research, that of the positivist paradigm and the interpretivist paradigm. Their
view on the most appropriate paradigm with regards to educational research is that of the
interpretivist, since it accommodates for the “immense complexity of the human nature”. I
agree that it is not wise to look at educational research in a purely positivist way which can
be cold and calculated, yet in relation to the nature of my particular research, it is the
primary paradigm that I will follow. Since the aims of this enquiry are ultimately to improve

50
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

mathematical ability, which is formally assessed by way of the summative GCSE

examinations, I feel that taking a predominantly positivist approach to the research enquiry
is most appropriate, with a secondary interpretivist element in order to triangulate my
results (Bell and Waters, 2014, Part 2).

Setting and Participants

I conducted my action research in School A, a larger than average-sized secondary school. It
accommodates the full range of abilities, with a special provision for those students with
more complicated learning needs. The school has a lower than average proportion of pupil
premium students, and less than a quarter of the participants involved in the research were
identified as fulfilling the pupil premium indicator.

The participant selection process was given much thought and planning in order to make
sure my study was feasible and that it would provide valid data. To begin, I made sure that I
had sufficient access to the participants that were to be involved in the instruction and
testing (Cohen, Manion and Morrison, 1993, p81-84). Being a large school, each year group
was split into two populations which each had different timetables. Having studied the data
of the classes to which I had already been assigned in School A, I searched for classes of the
same age and ability in opposing populations. In one of these instances, the two populations
had the same class teacher, and on speaking with the teacher, I established that their
abilities were very similar and that they would be suitable classes to use in my enquiry.
Throughout this enquiry, I have referred to Class A as the class who received the vocabulary
instruction via the Frayer Model, and Class B acted as the control group.

There were 15 students in each class who participated in the study, all of whom were of low
mathematical ability, and most of whom had a lower than average reading age, an indication
of vocabulary comprehension. Abedi and Lord (2001) found that there is a relatively greater
disadvantage for lower-ability students in answering questions that are composed of
complex language; therefore, I felt that the effects of the vocabulary instruction would be
highlighted to a greater extent with this low ability class, and they would potentially
experience the greatest improvement.

Design
Explicitly teaching keywords at the start of a lesson is the most common form of vocabulary
instruction that I have encountered. However, Monroe and Orme (2002) state that
definition-only teaching of vocabulary is ineffective and leads to minimal understanding, and
it is this surface level learning that I wanted to improve upon. Etymology is also a method
used by some teaching, yet as Morgan (1999) points out, with students rarely learning Latin
or Greek in today’s secondary schools, they are unlikely to be able to make the connection
with the word origins.

Some educationalists value learning through a vivid context that helps to motivate the
student (Boaler, 1994; Chapman, 2006), yet Monroe and Orme (2002) suggest that this
alone cannot provide the full depth of meaning. They conclude that a combined approach
has seen much better results.

The use of graphic organisers has been deemed to successfully fulfill this combined
approach of vocabulary instruction, if the teacher is able to facilitate their usage effectively.
There are many forms of the graphic organiser, such as the Concept of Definition developed
by Schwartz and Raphael (1985) that uses a semantic word map, and more recently, Rupley

51
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

et al. (2012) discussed the advantages of the Concept Wheel. The Frayer Model (Frayer,
Fredrick and Klausmeier, 1969) is a well-known example, yet not one that I have ever seen
used in the classroom. It has been modified from 13 vocabulary attributes to just 4, making
links between definitions, examples, non-examples and connections. These links are vitally
important for vocabulary comprehension, particularly the link to ‘connections’, which brings
in the background (existing) knowledge that provides the hooks allowing for retention and
recall of new knowledge (Marzano, 2004).

Graphic organisers have proven to be much more effective if presented as an activity at the
end of a period of direct instruction, context and utilisation (Moore and Readance, 1984);
this aids the organisation of the new word and, more importantly, the associated concept in
a way that best serves the retention and recall processes of the brain (Monroe and Orme,
2002).

Positive results are also seen if the graphic organisers are used as a discussion focus;
Vygotsky (1978) wrote extensively on the connection between language and development,
and a great deal has been written about the importance of discussion in learning new
vocabulary and concepts (Furner, Yahya and Duffy, 2005; Gough, 2007; Thompson and
Rubenstein, 2014).

However, Monroe and Pendergrass (1997) warned of using graphic organisers with low-
ability students as they often had less rich background knowledge with which to make the
conceptual connections. Conversely, Singleton and Filce (2015) have since noted that
graphic organisers have indeed proven to be a valuable tool in helping secondary students
who struggle to read and comprehend difficult texts. They make the critical distinction that
teachers must teach students how to use the graphic organisers and model them effectively,
harnessing whatever background knowledge they may have in order to build confidence, a
crucial point to which Monroe and Pendergrass (1997) failed to make any reference.

In light of this evidence, I planned to use the Frayer Model of the graphic organiser for
vocabulary instruction, rather than direct vocabulary instruction or instruction in context
alone. I built this activity into the end of the lessons, after a period of usage, as a way to
recapture the main elements of the new word. The activity was always completed as a class
activity; the mathematical ability of the class was low and many of the students also lacked
good literacy skills, so we discussed the elements of the model together and I combined the
ideas of the class in order to come up with definitions, examples and connections to
previous knowledge.

McConnell (2008) carried out similar action research that involved direct vocabulary
instruction. He concluded that the vocabulary instruction was valuable as it raised that test
scores of the participants. Yet the vocabulary test scores were compared with scores of the
same class in previous tests. Without further investigation, I could not be sure that it was the
vocabulary instruction alone that improved the scores; perhaps the topics being tested were
not so challenging for those students. To avoid this area of uncertainty, I decided to conduct
my research using two classes of very similar ability and in the same year group who were
completing the same units of work in parallel; one class received the usual form of
vocabulary teaching, while I would teach the other class using the Frayer Model, allowing
me to compare the progress made with and without the new style of instruction. I continued
with this instruction for a period of 6 weeks.

52
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

It was important that I first gained some baseline data in both classes that were to be
involved in the research in order to be able to assess any improvements, so I assessed their
current knowledge of the vocabulary in the forthcoming unit of work by way of a written
test prior to any instruction commencing. Students were asked to write a definition or
description of each word or give an example and non-example. Provided the students were
able to convey the concept of the mathematical word, I marked the question as correct. I did
not demand perfectly formed definitions at any stage due to the literacy abilities in the class;
plus there is never a need to provide a dictionary definition of a mathematical word in a
GCSE examination. I then re-assessed both classes at the end of the 6-week period using the
same vocabulary test.

The interpretivist element of my research approach took the form of observations

throughout the action research and results from a questionnaire. I briefly noted key
observations of the way that the students involved themselves with the graphic organiser
during the lessons, and then wrote the full details once each lesson had ended (Merriam,
2014, pp.128-131).

Participants who received the vocabulary instruction in the form of the Frayer Model were
also asked to complete a questionnaire at the end of the 6-week period. Due to the below
average literacy level of the class as a whole, I made sure that the language used in the
questionnaire was simple and statements were kept concise. I used a Likert scale to create
responses in the form of ‘Strongly agree’ to ‘Strongly disagree’ which I then coded in order
to analyse the data; however, I did not allow the respondents to see the codes as they can
be deemed as too confusing for children (Mellor and Moore, 2013). The questions were all
framed in a positive manner towards the method of vocabulary instruction, and therefore
the higher the score in the questionnaire, the more positively the participant engaged with
it. I aimed to avoid bias in the design of the questionnaire by using a 5-point balanced scale
with the same number of positive and negative options, and one neutral mid-point (Brace,
2008, pp. 67-73).

When collecting the data over the 6-week period, there were variables that were beyond my
control that I did not foresee. These included the absence of some students during some of
the lessons when I delivered the vocabulary instruction, which would have an impact of the
post-instruction test results. I believe that these absences were minimal however, so I have
not taken this into account when analysing the data, but have considered it a possible factor
when interpreting the results. Other variables included the motivation of the students to
engage with the model, and their ability to transfer their knowledge onto the test paper in
writing, both of which could be significant factors affecting the results of the study, yet
beyond the scope of this particular enquiry.

Data Analysis
I used Excel software in order to analyse the data from the test scores and the
questionnaire. Initially I represented this information by way of a dual bar chart to highlight
the differences between the two classes at a participant level.

The data from the vocabulary tests were then compared using averages (mean, mode and
median) and measures of spread (range, interquartile range, standard deviation and
variance) in relation to the change in scores; this allowed me to gain an understanding of the
scale of improvement made by each class. I have presented these differences visually by way
of box plots.

53
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

In order to help explain some of the differences in the test scores, I have analysed the
students’ reading ages compared to their improvement level in the vocabulary test using
regression analysis to determine if there is any correlation between the two variables.

Finally, I used the results of the questionnaire to help determine if there are other factors
that could affect the results of using the Frayer Model; for example, did those with low
levels of improvement record that they did not engage well this type of vocabulary
instruction?

Findings
At participant level, students in Class A made significantly greater gains, although some
students in both classes made no improvement at all.

Comparison of pre and post instruction vocabulary test

scores per participant - Class A
10

8
Test scores

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Participant

Pre Test Post Test

Figure 1. Bar chart showing the vocabulary test scores of Class A at the beginning and end of
the 6-week instructional period at participant level.

Comparison of pre and post instruction vocabulary test

scores per participant - Class B
10
8
Test scores

6
4
2
0
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Participant

Pre Test Post Test

Figure 2. Bar chart showing the vocabulary test scores of Class B at the beginning and end of
the 6-week instructional period at participant level.

54
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

Box plots to show the comparison of the

improvement in vocabulary test scores
between Class A and Class B

10
9
Change in vocabulary test score
8
7
6
5
4
3
2
1
0
Class A Class B

Comparing the results of the vocabulary test scores of both classes, Class A made the
greatest gains in vocabulary improvement, illustrated by the averages, although Class A’s
range of improvement was much broader than that of Class B.

Figure 3. Comparison of the improvement in vocabulary test scores over the 6-week
instructional period.

Table 1. A comparison of class averages and measures of spread based on the

improvements in vocabulary test scores.

Class A Class B
Mean 3.5 1.9
Median 3 2
Mode 3 2
Data points (n) 15 15
Minimum 0 0
Maximum 9 4
Range 9 4
Interquartile range 2.5 2
Variance 6.7 1.8
Standard Deviation 2.5 1.4

This amount of improvement in the vocabulary test scores appears to be related to the
participants’ reading ages; the product-moment correlation coefficient (r) is 0.55, indicating
a weak positive correlation.

55
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

Relationship between reading age and

improvement in vocabulary test scores.
10

Improvement in vocabulary test score

9
8
7
6
5
4
3
2
1
0
-40 -30 -20 -10 0 10 20 30
Difference between reading age and actual age (months)

Figure 4. Scatter graph with regression line showing the correlation between the difference
in reading age and actual age, and the improvement in vocabulary test scores.
Analysis of the questionnaire scores also shows a weak positive correlation with the
vocabulary test scores (product-moment correlation coefficient, r = 0.48).

Relationship between questionnaire score

and improvement in vocabulary test score.
30

25
Questionnaire score

20

15

10

0
0 2 4 6 8 10
Improvement in vocabulary test score

Figure 5. Scatter graph with regression line showing the relationship between the
improvement in the vocabulary test score and the questionnaire score.

Finally, when comparing the results of both classes in the unit assessment, there is an
insignificant difference in their scores, with Class A scoring a mean average of 54%
compared with 51% in Class B. However, only a third of the total marks available in the unit
assessment required knowledge of technical mathematical vocabulary; when analysing only
these marks, Class A scored 66% compared to Class B who scored 48%.

56
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

Discussion
Overall, the use of the Frayer Model as a method of vocabulary instruction appears to
improve the comprehension of technical vocabulary in mathematics, although greater gains
are generally seen in students who engaged well with the model and who had higher literacy
levels in terms of reading age. Participant 15, with a reading age of 26 months lower than
their actual age, did not make any improvements between their pre- and post-instruction
vocabulary test results, and scored only 6 on the questionnaire; additionally they
commented that the model was not enjoyable, despite efforts to teach the students how to
use and engage with the model, as recommended by Singleton and Filce (2015). At the other
end of the spectrum, participant 3 has a reading age of 21 months higher than their actual
age, and made an improvement of 6 marks on the vocabulary test; plus they recorded a high
score of 23 in the questionnaire.

This may indicate that reading age has some impact on the effectiveness of this method.
Indeed, Stanovich (2008) writes about the ‘Matthew effect’ whereby stronger reading
comprehension leads to stronger vocabulary knowledge and vice versa. Marzano (2004)
suggests that good vocabulary knowledge is due to a rich background knowledge, which
would explain why some students struggled with completing the ‘connections’ section of the
model, arguably the most important element for effective long-term retention and recall.
This result is in line with the writings of Monroe and Pendergrass (1997) who warn that
background knowledge is a success factor.

When considering the performance of the participants during the unit assessment, it was
encouraging to see that the students in Class A scored significantly higher in vocabulary-rich
questions than Class B, signalling that this purposeful vocabulary instruction has indeed lead
to gains in academic achievement, although perhaps at the expense of learning other
mathematical skills, shown by very similar overall assessment results between the two
classes.

An interesting observation occurred when studying the definitions that participants gave of
the mathematical word ‘product’; prior to the vocabulary instruction, some students in Class
A believed this word to be related to items that they would buy from a shop, and falls into
the category of everyday words that take a different mathematical meaning, as Gough
(2007) discusses. The interesting result from the post-instruction test showed that, even
though the correct mathematical definition of ‘product’ was taught using the Frayer Model,
some students still reverted back to their original, everyday definition in the test. This
confirms that definitions in this subcategory of mathematical words do indeed cause
problems for some students, and that the Frayer Model may not be the right tool to use for
such vocabulary, an area of investigation worth researching in more depth in the future.

The broad range of improvement in class A may indicate that this is a more efficient and
effective method of learning mathematical vocabulary for some students, yet it is not
advantageous for all students, highlighting the fact that teaching needs to be as personalised
as feasibly possible so as to enable all students to make their greatest potential progress.

Further Investigation
Although the result show a positive reflection of the model, further investigation would be
needed to confirm that it is this model in particular that has provided the benefit, and not
just the purposeful and increased time spent on vocabulary instruction. Some of the
participants commented that they preferred using mind maps to learn vocabulary, finding
them easier to produce and more helpful in recalling information, the positive aspect of this

57
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

reflection being that mind maps too are a form of graphic organiser. In which case it may be
conclusive that graphic organisers in general are beneficial, and students should be given the
choice of which to use.

During my observations of students completing the Frayer model, I noticed that I could
accurately use the model and the associated questioning, responses and general discussion
as a method of formative assessment. Upon marking the students’ work, I was also able to
ascertain which students understood the concepts behind the vocabulary, as there was a
marked difference in completed models, despite discussing the new terminology as a class;
it may be worth considering their use as plenaries or exit cards in future.

Limitations
Many researchers describe the importance of discussion in using this model (Furner, Yahya
and Duffy, 2005; Gough, 2007; Thompson and Rubenstein, 2014). In delivering this
vocabulary instruction, I was aware that it was difficult to find enough time during lessons to
implement it with the appropriate level of discussion required without sacrificing time used
to practise the skills content. Indeed, one of the participants commented that they would
have liked more time during lessons to discuss and complete the vocabulary model.
However, with the new GCSE bringing a significant increase in content (Department for
Education, 2013a), finding more time to teach vocabulary in maths lessons is increasingly
difficult.

Another limitation of this study involves the validity of the reading ages; these were
determined 4 months before I conducted the study, so may not have provided completely
accurate reflections of reading comprehension ability of the participants. I endeavoured to
mitigate any inaccuracies by comparing their actual age at the time the reading age was
determined.

Conclusion
Vocabulary knowledge is a success factor in the mathematical achievement of students of all
abilities (Aiken, 1972; Morgan, 1999; Schleppegrell, 2007). The language of mathematics has
a complex structure with technically specific vocabulary that can form a barrier to
achievement in lessons and assessments (Chinn, 2012; Gough, 2007), which is more of any
issue now that the new GCSE tests problem solving to a greater extent (Department for
Education, 2013a).

There is much evidence to show that the use of graphic organisers is a successful way to
represent the elements of new mathematical vocabulary (Frayer, Fredrick and Klausmeier,
1969; Singleton and Filce, 2015), although Monroe and Pendergrass (1997) warn of their
limitations with low ability students who may not have such a rich background knowledge.

The findings of this action research indicate that there is evidence to suggest that the Frayer
Model of graphic organiser can have more success in a low-ability class than direct
vocabulary instruction alone, as received by the control group. However, it may not be a
method that all students engage with, and there appears to be a weak correlation here with
a student’s reading age, an indication of reading comprehension ability. Importantly, the
findings confirm that this is not a method of vocabulary instruction that suits all students,
and the teaching of vocabulary, like all other aspects of teaching, should be personalised as
much as is feasibly possible.

58
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

An unexpected finding of the use of this model is its potential as a method of formative
assessment. However, the model needs sufficient time to be completed, a resource that is in
increasingly short supply in the mathematics classroom.

Reference List
Abedi, J. and Lord, C. (2001) ‘The Language Factor in Mathematics Tests’, Applied
Measurement in Education, 14(3), pp. 219-234. doi: 10.1207/S15324818AME1403_2
(Accessed: 28 November 2015)
Adams, TL (2003) ‘Reading mathematics: More than words can say', Reading Teacher, 56, 8,
p. 786, Academic Search Complete, EBSCOhost (Accessed 5 April 2016)
Aiken, LR (1972) Language Factors in Learning Mathematics, Available at:
http://files.eric.ed.gov/fulltext/ED068340.pdf, (Accessed 31 March 2016)
Amen, J. (2006) ‘Using Math Vocabulary Building to Increase Problem Solving Abilities in a 5th
Grade Classroom’, Action Research Projects, Paper 13, Available at
http://digitalcommons.unl.edu/mathmidactionresearch/13. (Accessed 30 December
2015)
Barker Bausell, R. (2015) The Design and Conduct of Meaningful Experiments Involving
Human Participants. Oxford: Oxford University Press
Bell, J. and Waters, S. (2014) Doing Your Research Project : A Guide For First-Time
Researchers, Maidenhead: McGraw-Hill Education, eBook Collection (EBSCOhost),
EBSCOhost (Accessed 5 April 2016)
Biggs, J.B. and Collis, J.F. (1982) Evaluating the quality of learning – the SOLO taxonomy (1st
ed) New York: Academic Press
Bloom, B.S. (1956) Taxonomy of Educational Objectives: The Classification of Educational
Goals. London: Longman Group Ltd
Boaler, J. (1994) 'When do girls prefer football to fashion? An analysis of female in
underachievement in relation to real-life mathematical contexts', British Educational
Research Journal, 20(5) p. 551, Academic Search Premier, EBSCOhost (Accessed: 12
December 2015)
Brace, I (2008) Questionnaire Design : How to Plan, Structure and Write Survey Material for
Effective Market Research (2nd Edition), Kogan Page Ltd., London, GBR. Available
from: ProQuest ebrary. (Accessed 3 March 2016)
Braselton, S. and Decker, B (1994) 'Using graphic organizers to improve the reading of
mathematics', Reading Teacher, 48, 3, p. 276, Academic Search Complete, EBSCOhost,
(Accessed 31 March 2016)
Carter, T.A. and Dean, E.O. (2006) ‘Mathematics Intervention for Grades 5–11: Teaching
Mathematics, Reading, or Both?’, Reading Psychology, 27(2-3), pp. 127-146, doi:
10.1080/02702710600640248 (Accessed: 28 November 2015)
Chapman, O. (2006) 'Classroom Practices for Context of Mathematics Word Problems',
Educational Studies In Mathematics, 62(2) pp. 211-230, Academic Search Premier,
EBSCOhost (Accessed: 12 December 2015)
Chinn, S. (2012) The Trouble with Maths – A practical guide to helping learners with
numeracy difficulties. (2nd Edn) Abingdon, Oxon: Routledge
Cohen, L., Manion, L. and Morrison, K. (1993) Research Methods in Education. Florence, GB:
Routledge, ProQuest ebrary. (Accessed 6 April 2016)
Department for Education (2013a) Reformed GCSEs in English and mathematics. School and
college qualifications and curriculum. Available at
https://www.gov.uk/government/speeches/reformed-gcses-in-english-and-
mathematics (Accessed: 29 December 2015)
Department for Education (2013b) Mathematics GCSE Subject content and assessment
objectives. Available at: https://www.gov.uk/government/uploads/system/uploads/

59
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

attachment_data/file/254441/GCSE_mathematics_subject_content_and_assessment_objec
tives.pdf (Accessed 23 March 2016)
Elton-Chalcraft, S., Hansen, A. and Twiselton, S. (2008) Doing Classroom Research: A Step-by-
Step Guide for Student Teachers. Maidenhead: Open University Press
Frayer, D.A., Fredrick, W.C. and Klausmeier, H.J. (1969) A Schema For Testing the Level of
Concept Mastery, Working Paper No.16, Available at:
http://brainimaging.waisman.wisc.edu/~perlman/frayer-frederick-klausmeier.pdf
(Accessed: 2 January 2016)
Furner, J., Yahya, N., and Duffy, M. (2005) 'Teach Mathematics: Strategies to Reach All
Students', Intervention In School & Clinic, 41(1), pp. 16-23, Academic Search Premier,
EBSCOhost (Accessed 28 November 2015)
Gough, J. (2007) 'Conceptual complexity and apparent contradictions in mathematics
language', Australian Mathematics Teacher, 63(2) pp. 8-16, Academic Search Premier,
EBSCOhost (Accessed: 28 November 2015)
Kan, A. and Bulut, O. (2015) 'Examining the Language Factor in Mathematics Assessments',
Education Research & Perspectives, 42, 1, pp. 582-606, Education Source, EBSCOhost,
(Accessed 5 April 2016)
Marzano, R.J. (2004) Building Background Knowledge For Academic Achievement: Research
on What Works in Schools. Alexandria, VA: Association for Supervision and Curriculum
Development. eBook Collection (EBSCOhost), EBSCOhost (Accessed: 13 December
2015)
McConnell, M. (2008) "Exploring the Influence of Vocabulary Instruction on Students’
Understanding of Mathematical Concepts" (2008). Action Research Projects. Paper 54.
Available at: http://digitalcommons.unl.edu/mathmidactionresearch/54 (Accessed 4
April 2016)
Mellor, D. and Moore, K (2014) 'The Use of Likert Scales With Children', Journal Of Pediatric
Psychology, 39, 3, pp. 369-379, Education Source, EBSCOhost (Accessed 7 March
2016)
Merriam, Sharan B. (2014) Qualitative Research: A Guide to Design and Implementation (3).
Somerset, US: Jossey-Bass, ProQuest ebrary. (Accessed 15 Jan 2016)
Monroe, E.E. and Orme, M.P. (2002) ‘Developing Mathematical Vocabulary’ Preventing
School Failure: Alternative Education for Children and Youth, 46(3), pp.139-142, doi:
10.1080/10459880209603359 (Accessed: 3 December 2015)
Monroe, E.E. and Pendergrass, M. (1997) ‘Effects of Mathematical Vocabulary on fourth
grade students’, Reading Improvement, 34(3), pp120-132. Available at
http://files.eric.ed.gov/fulltext/ED414182.pdf (Accessed: 6 January 2016)
Morgan, C. (1999) ‘Communicating Mathematically’, In Johnston-Wilder, S., Johnston-
Wilder, P., Pimm, D. and Westwell, J. eds. Learning to Teach Mathematics in the
Secondary School. London: Routledge
Norman, G (2010) 'Likert scales, levels of measurement and the "laws" of statistics',
Advances In Health Sciences Education: Theory And Practice, 15, 5, pp. 625-632,
MEDLINE with Full Text, EBSCOhost, (Accessed 31 March 2016)
Ofqual (2015) A comparison of actual and expected difficulty, and assessment of problem
solving in GCSE maths. Available at https://www.gov.uk/government/uploads/
system/uploads/attachment_data/file/429117/2015-05-21-gcse-maths-research-on-sample-
assessment-materials.pdf (Accessed 04 April 2016)
Riccomini, P.J., Smith, G.W., Hughes, E.M. and Fries, K.M. (2015) ‘The Language of
Mathematics: The Importance of Teaching and Learning Mathematical Vocabulary’,
Reading & Writing Quarterly, 31(3), pp. 235-252, doi:
10.1080/10573569.2015.1030995 (Accessed 22 November 2015)

60
 KENYON: HOW CAN WE IMPROVE MATHEMATICAL VOCABULARY COMPREHENSION THAT
WILL ALLOW STUDENTS TO DEVELOP HIGHER-ORDER LEVELS OF LEARNING?

Rupley, W., Dee Nichols, W., Mraz, M., and Blair, T. (2012) 'Building Conceptual
Understanding through Vocabulary Instruction', Reading Horizons, 51(4) pp. 299-320,
Education Source, EBSCOhost (Accessed: 12 December 2015)
Schleppegrell, M.J. (2007) ‘The Linguistic Challenges of Mathematics Teaching and Learning:
A Research Review’, Reading & Writing Quarterly, 23:2, 139-159, DOI:
10.1080/10573560601158461 (Accessed 5 April 2016)
Schwartz, R. and Raphael, T. (1985), ‘Concept of Definition: A key to improving students’
vocabulary’, International Reading Association, The Reading Teacher, 39(2) pp. 198 –
205. Available at
https://edc448uri.wikispaces.com/file/view/Schwartz%20%26%20Raphael%
201985.pdf/94665074/Schwartz%20%26%20Raphael%201985.pdf. (Accessed 1 January
2016)
Stanovich, K.E. (2008) ‘Matthew Effects in Reading: Some Consequences of Individual
Differences in the Acquisition of Literacy’, Journal of Education, 189(1/2), pp.23-55.
Academic Search Premier, EBSCOhost (Accessed: 13 December 2015)
Thompson, D. and Rubenstein, R. (2014) 'Literacy in Language and Mathematics', Journal Of
Adolescent & Adult Literacy, 58(2) pp. 105-108, Academic Search Premier, EBSCOhost
(Accessed: 22 November 2015)
Vygotsky, L. (1986) Thought and Language. Cambridge, MA: Massachusetts Institute of
Technology
Wakefield DV (2000) ‘Math as a Second Language’, The Educational Forum, 64:3, 272-279,
DOI: 10.1080/00131720008984764 (Accessed 04 April 2016)
Wieher, BM (2010) 'Vocabulary words count in math, too', Reading Today, 28, 3, p. 45,
Academic Search Complete, EBSCOhost (Accessed 1 April 2016)
Zeni, J. (1998) ‘A guide to ethical issues and action research’, Educational Action Research,
6:1, 919, DOI: 10.1080/09650799800200053 (Accessed 5 April 2016)

61